JP2000308046A - Data generation method and data transmission device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a data transmission device where an end code can easily be added by adding invalid data to compressed video data and inserting a code showing a break into compressed video data in a fixed length as compressed data of the fixed length. SOLUTION: A video data storage area is installed in a part of an SDRAM part 250 being an outer memory. Invalid data are previously written into all the video data storage areas. At the time of reading compressed video data from the SDRAM part 250, all data in the storage area is read from the head of the video data storage area. Compressed video data in a variable length becomes compressed video data in the fixed length by the reading. When all data are read and they become the number of words for one frame from a word counting part 214, and end code generated in an SDTI core part 222 is added to final data. Thus, data management for adding the end code is reduced in a memory management part 210.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a method for generating compressed video data and a data transmission apparatus. More specifically, by inserting stuffing data, which is invalid data, into fixed-length data by compressing one frame or one field of variable-length video data, inserting a code indicating a break of variable-length video data. Can be easily performed. The present invention also relates to a data transmission device provided with a memory for storing such fixed-length compressed video data.

[0002]

2. Description of the Related Art As a transmission format for transmitting digital data, SMPTE-standardized by SMPTE is used.
259M "10-bit 4: 2: 2 Component
ntand 4fsc Composite Digit
al Signals-Serial Digita
l Interface ”(hereinafter“ SDI (Serial Digi
tal Interface) format) is known. This transmission standard is a standard for transmitting a digital component video signal (D1 signal), a digital composite signal (D2 signal), and the like.

As is well known in this transmission format,
Although it is a standard that allows audio data and other data to be inserted into the ancillary data area and transmitted, the format is not such that compressed video data is inserted into the active video area and transmitted. This is because the amount of encoding of the compressed data varies depending on the line.

[0004] The transmission format is standardized so that even such compressed data can be transmitted.
05M "Serial Data Transport"
Interface "(hereinafter referred to as" SDTI format "). That is, this is the SDT corresponding to the active video area ACV in the SDI format.
Multi-data such as compressed video data and uncompressed audio data is inserted into a payload section on the I format so that the data can be transmitted.

[0005] When transmitting video data, uncompressed audio data, and the like (hereinafter also referred to as effective data) compressed using the SDTI format, predetermined additional data is added to the effective data, and data in the SDTI format is added. You need to create a stream.

FIG. 15 is a block diagram showing an outline of such a data transmission apparatus. A terminal 12 has compressed video data such as MPEG (Moving Picture Experts G).
roup) The video data DVC compressed according to the standard is supplied, the uncompressed audio data is supplied to the terminal 14, and the data other than the effective data is supplied to the terminal 16.

The video data DVC is transmitted to the interface 22
Is supplied to the memory management unit 30 via the external memory.
0 is supplied. Similarly, audio data DAU, AU
The X data DSX is also stored in the memory management unit 30
Is supplied to the SDRAM 840 through the memory and stored.

Here, the video data DVC is, for example, 27M
It is supplied as 8-bit data with a transmission rate of Hz, 36 MHz or 54 MHz, and 16-bit audio data DAU with a sampling frequency of 48 kHz.

Since the transmission speed and the sampling frequency of the data supplied from the outside are different from each other, the interfaces 22, 24, and 26 switch the data input at different frequencies to the clock frequency data of the SDRAM unit 40. Is also performed.

The SDRAM unit 40 is written with a fixed amount of data as a unit, read out with a fixed amount of data as a unit, and supplied to the formatter 50. In the formatter 50, a process of adding additional data for the SDTI format to the above-described effective data is performed.

That is, "separ" is added to the variable-length video data.
attor "," wood count "," end c "
ord ”is added. An end synchronization code EAV, ancillary data ANC, a start synchronization code SAV, and the like are added before the payload portion. The data in the SDTI format is converted into serial data by the parallel / serial converter 60 and then transmitted to a transmission line (not shown) at a transmission speed of 270 Mbps or the like.

Interfaces 22, 24, 26, memory management unit 30, formatter 50 and P / S
The converter 60 supplies the necessary clock frequency (27 MHz / 81 MHz / 108 MH) from the clock generator 70.
z) is provided.

[0013]

By the way, compressed video data is variable-length data, and the amount of compressed data differs depending on its frame (or field). Therefore, to separate the compressed data in frame units, as described above, “Sepa” is added to the beginning of the variable-length compressed video data.
rat ", and" End Co. "
de "is added. This process is performed by the memory management unit 30.

However, whether or not the read compressed video data is the last data cannot be easily grasped because the compressed video data is variable-length data.
Therefore, when supplying the compressed video data to the SDRAM unit 40, the number of words for one frame is counted, and the last word of the read compressed video data is detected while referring to the stored word count number. I have.
When it is determined that this is the last word, an end code is added.

As described above, in the related art, the process of adding the end code is unexpectedly troublesome, and the memory management unit 30 is burdened accordingly.

Therefore, the present invention has been made to solve such a conventional problem, and the end code can be easily added by treating the variable-length compressed video data as fixed-length compressed video data. A data transmission device is proposed.

[0017]

According to a first aspect of the present invention, there is provided a data generating method according to the present invention, wherein one frame or one field of variable-length compressed video data is invalid data. Is added as a fixed-length compressed data, and a code indicating a break of one frame or one field of compressed video data is inserted into the fixed-length compressed video data. .

Further, in the data transmission apparatus according to the present invention, the section of one line of the video frame is defined as:
An end synchronization code area in which an end synchronization code is inserted, an auxiliary data area in which auxiliary data is inserted, a start synchronization code area in which a start synchronization code is inserted, and a payload area in which data including compressed video data is inserted. Wherein the compressed video data inserted into the payload section is converted into one frame or one field of variable-length compressed video data. The compressed video data is fixed-length compressed video data to which invalid data is added.

According to the present invention, the SDRA which is an external memory
A video data supply area is provided in the M section. This area is a storage area in which data is written in frame units, and its data length (the number of words) is fixed. Then, invalid data (stuffing data) “0” is previously written in the fixed-length video data storage area (data supply area),
Write (overwrite) compressed video data from its head area.

The compressed video data is variable-length data that differs depending on the video content of the frame. When the writing of the compressed video data is completed, the data “0” remains as it is because the remaining area is not overwritten.

When data is read, all data (for a fixed number of words) of the storage area is read from the head of the video data storage area. Thus, the variable-length compressed video data becomes fixed-length compressed video data. After reading all data, an end code is added. Since the number of read words is constant in any frame, the management of the number of read words is easy. As a result, data management for adding an end code in the memory management unit is reduced.

[0022]

Embodiments of the present invention will be described below with reference to the drawings. In the present embodiment, data such as video and audio materials are packaged and each content item (for example, a picture item (Picture Item) or an audio item (Audio Item)) is packaged.
m)), as well as packaging information about each content item, metadata about each content, and the like to form one content item (system item (S
ystem Item)), and make each of these content items a content package. Further, a transmission packet is generated from the content package and transmitted using the serial digital transfer interface.

As the serial digital transfer interface, for example, SMPTE-259M "10-bit 4: 2: 2" standardized by SMPTE
Component and 4fsc Composi
te Digital Signals -Seria
l Digital Interface ”(hereinafter“ Serial Digital Interface SDI (Serial Digita
l Interface) format) and the SMPTE-305M “Seri” standard for transmitting packetized digital signals.
al DataTransport Interface
e "(hereinafter referred to as" SDTI format ") to transmit the above-mentioned content package.

First, when an SDI format standardized by SMPTE-259M is arranged in a video frame, an NTSC 525 digital video signal is
1716 (4 + 268 + 4 +) per line in the horizontal direction
1440) The word is composed of 525 lines in the vertical direction. Also, a digital video signal of the PAL625 system is 1728 (4 + 28) per line in the horizontal direction.
(0 + 4 + 1440) words and 625 lines in the vertical direction. However, it is 10 bits / word.

For each line, four words from the first word to the fourth word are used as a video signal area 144.
Indicates the end of the 0-word active video area, and is used as an area for storing a code EAV (End of Active Video) for separating the active video area and an ancillary data area described later.

For each line, 268 words from the fifth word to the 272th word are used as an ancillary data area, and header information and the like are stored. Four words from the 273rd word to the 276th word indicate the start of the active video area, and a code SA for separating the active video area from the ancillary data area.
It is used as an area for storing V (Start of Active Video), and the 277th word and subsequent words are used as an active video area.

In the SDTI format, the above-mentioned active video area is used as a payload area, and the codes EAV and SAV indicate the end and start of the payload area.

Here, the data of each item is inserted into the payload area of the SDTI format as a content package, and the code EAV of the SDI format is inserted.
And SAV are added to form data in a format as shown in FIG. When data in the format shown in FIG. 1 (hereinafter referred to as "SDTI-CP format") is transmitted, P / S conversion and transmission path encoding are performed and transmitted as serial data, as in the SDI format and SDTI format. You. In FIG. 1, the numbers in parentheses indicate the numerical values of the PAL625 video signal, and the numbers without the parentheses indicate the numerical values of the NTSC 525 video signal. Hereinafter, only the NTSC system will be described.

FIG. 2 shows the structure of the code EAV and the header data included in the ancillary data area.

The code EAV is 3FFh, 000h, 000
h, XYZh (h indicates hexadecimal notation, and the same applies to the following description).

In "XYZh", the bit b9 is set to "1" and the bits b0 and b1 are set to "0". Bit b8 is a flag indicating whether the field is the first field or the second field.
Is a flag indicating a vertical blanking period. Bit b6 is a flag indicating whether the 4-word data is EAV or SAV. The flag of the bit b6 is set to "1" for EAV and "0" for SAV. Bits b5 to b2 are data for performing error detection and correction.

Next, at the beginning of the header data, data "ADF (Ancillary data flag)" for header data recognition is used.
Are fixed patterns 000h, 3FFh, 3FFh. Following this fixed pattern, "DID (Dat
a ID) "and" SDID (Secondary data ID) ", and fixed patterns 140h and 101h are arranged in the SDTI-CP format.

"Data Count" is replaced with "Line
Number-0 ”to“ Header CRC1 ”
This indicates the number of words up to 46 words, and the number of words is 46 words (22Eh).

"Line Number-0, Line
"Number-1" indicates the line number of the video frame. In the NTSC 525 system, these two words indicate the line numbers from 1 to 525. In the PAL system 625 system, line numbers from 1 to 625 are indicated.

"Line Number-0, Line
Number-1 followed by Line Number
er CRC0, Line Number CRC1 "
Are arranged, and this “Line Number CR”
C0, Line Number CRC1 ”is“ DI
D ”to“ Line Number-1 ”, CRC (cyclic redundancy check) for 5-word data
codes), which are used for checking transmission errors.

"Code & AAI (Authorized addr
ess identifier) indicates information such as what the word length of the payload area from the SAV to the EAV is set and what data format is used for the addresses of the sending side and the receiving side. .

[Destination Addresses]
"s" is the address of the data receiving side (transmission destination), and "S
"source Address" is the address of the data transmission side (transmission source).

The “Block Type” following the “Source Address” indicates the format of the payload area, for example, whether the payload area has a fixed length or a variable length. Compressed data is inserted. Here, SDTI-C
In the P format, for example, when a content item is generated using compressed video data (video data), a variable length block (Variable Block) is used because the data amount differs for each picture. For this reason, SD
"Block Type" in TI-CP format
Is fixed data 1C1h.

"CRC Flag" indicates whether a CRC is placed in the last two words of the payload area.

Also, "Da" following "CRC Flag"
“ta extension flag” indicates whether or not the user data packet is extended.

"Data extension fla
Following "g", a 4-word "Reserved" area is provided. Next "Header CRC 0, Head
erCRC 1 ”is a CRC for data from“ Code & AAI ”to“ Reserved 4 ”.
(Cyclic redundancy check codes), which are used for checking transmission errors. The next "Check Su
“m” is a Check Sum code for all header data, and is used for checking transmission errors.

In the payload area shown in FIG. 1, data of items such as video and audio is packaged as a variable-length block in SDTI format. FIG. 3 shows the format of a variable-length block. "Separator" and "End Cod"
"e" indicates the start and end of the variable-length block,
The value of “Separator” is “309h”, and “End”
The value of “Code” is set to “30 Ah”.

The “Data Type” indicates what kind of item the packaged data is. The value of the “Data Type” is, for example, “04h” in the system item (System Item), "05h" for picture items, "06h" for audio items, and "07h" for AUX items that are other data.
It is said. Note that one word is 10 bits as described above. For example, when the word is 8 bits as indicated by “04h”, the 8 bits correspond to bits b7 to b0. Further, by adding even parity of bits b7 to b0 as bit b8 and adding logically inverted data of bit b8 as bit b9, data of 10 bits is obtained. The 8-bit data in the following description is similarly converted into 10 bits.

In "Word Count", "Data
The “Dat” indicates the number of words of “Block”.
"a Block" is the data of each item. Here, the data of each item is packaged in picture units, for example, frame units, and in the NTSC system, the program switching position is set to the position of 10 lines. Therefore, in the NTSC system, as shown in FIG. 1 in
From the third line, system items, picture items,
Audio items and AUX items are transmitted in this order.

FIG. 4 shows the structure of a system item. "System ItemType" and "Wor
“d Count” is “Data Ty” of the variable-length block.
pe "and" Word Count ".

The one-word "System Item B"
The bit b7 of "itmap" is a flag indicating whether or not an error detection and correction code such as a Reed-Solomon code is added, and when "1" is set, the error detection and correction code is added. Is shown. Bit b6 is a flag indicating whether or not there is SMPTE Label information. Here, when “1” is set, it indicates that the information of the SMPTE Label is included in the system item. Bits b5 and b
4 is Reference Date / Time status
mp, Current Date / Time stamp
This is a flag indicating whether or not p is in the system item. This Reference Date / Times
In the “tamp”, for example, the time or date when the content package was first created is indicated. Also Curren
In t Date / Time stamp, the time or date when data of the content package was last corrected is indicated.

Bit b3 is a picture item, bit b
2 is an audio item, and bit b1 is a flag indicating whether or not the AUX item is after the system item. When "1" is set, it indicates that the item exists after the system item.

Bit b0 is the control element (C
ontrol Element).
When "1" is set, it indicates that the control element exists. In addition, although not shown, the bit b
8, b9 are added as described above and transmitted as 10-bit data.

One word "Content Packa"
bits b7 to b6 of "ge Rate" are undefined areas (Res
erved), and the bits b5 to b1 indicate the package rate which is the number of packages per second in the 1 × speed operation.
(Package Rate) is displayed. Bit b0 is a 1.001 flag, and when the flag is set to "1", it indicates, for example, a 59.94 (= 60 / 1.001) field per second.

One word "Content Packa"
Bits b7 to b5 of “ge Type” are “Stream States” flags for identifying the position of the picture unit in the stream. This 3
The following eight types of states are indicated by the bit flags.

0: The picture unit is a pre-roll (pr
e-roll) section, editing section, post-roll
It does not belong to any of the sections. 1: This picture unit is a picture included in the pre-roll section, and is followed by an editing section. 2: This picture unit is the first picture unit of the editing section. 3: This picture unit is a picture unit included in the middle of the editing section. 4: This picture unit is the last picture unit of the editing section. 5: This picture unit is a picture unit included in the post-roll section. 6: This picture unit is the first and last picture unit of the editing section (the state where the editing section has only one picture unit). 7: Not defined

The bit b4 is an undefined area (Reserved), and the bits b3 and b2 of "Transfer Mod"
"e" indicates the transmission mode of the transmission packet. Also, the transmission timing mode when transmitting the transmission packet is indicated by “Timing Mode” of bits b1 and b0. Here, when the value indicated by the bits b3 and b2 is “0”, the synchronous mode (Synchronous mode)
When "1", isochronous mode,
When it is "2", it is set to an asynchronous mode (Asynchronous mode). When the value indicated by the bits b1 and b0 is “0”, the transmission of the content package for one frame is started at a timing of a predetermined line of the first field (Normal timing mode).
Advanced timing mode in which transmission is started at the timing of a predetermined line in the second field when "1"
(Advanced timing mode), when "2", a dual timing mode in which transmission is started at a timing of a predetermined line in each of the first and second fields (Dual timing mode)
mode).

"Content Package Ty"
"Channel" followed by two words "Channel Handl"
"e" is used to determine the content package of each program when the content packages of a plurality of programs are multiplexed and transmitted, and are identified by identifying the values of bits H15 to H0. Content packages can be separated for each program.

The two-word “Continuity Co”
"unt" is a 16-bit modulo counter. This counter is counted up for each picture unit, and is independently counted for each stream. Therefore, when a stream is switched by a stream switcher or the like, the value of this counter becomes discontinuous, and a switching point (editing point) can be detected. Since this counter is a 16-bit modulo counter as described above and has a very large value of 65536, the probability that the value of the counter coincidentally coincides at the switching point between the two switched streams is extremely low. In addition, practically sufficient accuracy can be provided for detecting the switching point.

After “Continuity Count”, the above-mentioned SMPTE Label or Refer
ence Date / Time and Current
"SMPTE Universal" indicating Date / Time
sal Label "," Reference Dat "
e / Time stamp "," Current Da
A "te / Time stamp" area is provided.

After that, "Package Meta"
data Set ”or“ Picture Metada ”
ta Set "," Audio Metadata Se "
t "" Auxiliary Metadata Se
A "t" region is provided. In addition, "Picture Me
tadata Set "," Audio Metadat "
a Set "," Auxiliary Metadata "
Set ”indicates that the corresponding item is“ System I
It is provided when it is indicated that the content is included in the content package by the flag of “tem Bitmap”.

In the above-mentioned “Time stamp”, 17 bytes are allocated, and “Time” is used in the first byte.
e stamp ", and the remaining 16 bytes are used as a data area. here,
The first 8 bytes of the data area are, for example, SMPTE12
Time code (Timecode) standardized as M is shown, and the subsequent 8 bytes are invalid data.

As shown in FIG. 5, the time code of 8 bytes is “Frame”, “Seconds”, “Minute”.
s "," Hours "and 4-byte" Binary "
Group Data ”.

Bits b5 and b4 of "Frame" indicate the tens place of the frame number, and bits b3 to b0 indicate the place of the first place. Similarly, “Secons” and “Minute”
s ”,“ Hours ”, each bit b6 to b0 indicates seconds,
Minutes and hours are shown.

Bit b7 of “Frame” is a color frame flag (Color Frame Flag), and indicates whether the frame is the first color frame or the second color frame. Bit b6 is a drop frame flag (Drop Frame flag).
Flag), which is a flag indicating whether or not the video frame inserted in the picture item is a drop frame. Bit b7 of “Seconds” is, for example, NTSC
In the case of the system, a field phase (Field Phase), that is, whether the field is the first field or the second field is indicated. In the case of the PAL method, the field phase is indicated by the bit b6 of “Hours”.

Three bits B0 of bit b7 of "Minutes" and bits b7 and b6 of "Hours"
B3 (In the PAL system, “Seconds” and “Mini
tes ”and“ Hours ”, each bit b7) indicates whether or not there is data in each of BG1 to BG8 of“ Binary Group Data ”. In this “Binary Group Data”, for example, the Gregorian calendar (GregorianCalender) and the Julian calendar (Julian calendar)
Calender) can be displayed in two digits.

FIG. 6 shows the structure of "Metadata Set", in which one word "Metadata Co" is used.
"metadata Bl" in the set
The number of “ocks” is indicated. In addition, "Metadata
When the value of “Set” is 00h, “Metadata
Since there is no "Block", "Me
“tadata Set” is one word.

Here, “Metadata Bloc
“k” indicates “Package Metadata Set” indicating content package information such as a program title.
In the case of, the one-word "Metadata Type
"e", a two-word "Word Count", and an information area "Metadata" is provided.
If the number of words of this “Metadata” is “Word C
"out" is indicated by bits b15 to b0.

“Picture Metadata Set”, “A” indicating information related to packaged items such as video, audio or AUX data
audio Metadata Set "and" Auxili
The “ary MetadataSet” further includes “Element Type” and “Element” of one word.
“Element Number” is provided, and “Element Data” of items such as video and audio described below is provided.
"Element Type" and "E"
element Number ", and is linked to" Element Data Block ".
Metadata can be set for each time. After these “Metadata Set”, “Cont.
A "rollElement" region can be provided. Next, a block of each item such as video and audio will be described with reference to FIG. The block “Item Type” of each item such as video and audio indicates the type of the item as described above, “05h” for a picture item, “06h” for an audio item,
In the AUX data item, "07h" is set. "It
In the "em Word Count", the number of words up to the end of this block ("Word Co
unt ”). "Item Word
"Item Header" following "Count" indicates the number of "Element Data Block". Here, since “Item Header” is 8 bits, “Element DataBlo” is used.
The number of “ck” ranges from 1 to 255 (0 is invalid). "Element" following this "Item Header"
“Data Block” is the data area of the item.

"Element Data Bloc"
k "means" Element Type "and" Element
t Word Count "," Element Num "
ber "and" Element Data ", and" Element Type "and" Element ".
Word Count "by" Element
The type and amount of data of “Data” are shown.
The “Element Number” indicates the number of the “Element Data Block”.

Next, the structure of “Element Data” will be described. MPEG, one of the elements
The -2 picture element is an MPEG-2 video elementary stream (V-ES) of any profile or level. Profiles and levels are defined in the decoder template document. FIG. 8 is an example of the format of MPEG-2 V-ES in the SDTI-CP element frame. This example is a V-ES bitstream that specifies a key, ie, an MPEG-2 start code (according to SMPTE recommended practices). MPEG
The -2 V-ES bitstream is simply formatted into data blocks as shown in FIG.

Next, metadata for a picture item, for example, MPEG-2 picture image editing metadata will be described. This metadata is a combination of edit and error metadata, compression encoded metadata, and source encoded metadata. These metadata can be inserted mainly in the system items described above, as well as in auxiliary data items.

FIG. 9 shows “Picture Editing Bitma” provided in the MPEG-2 picture editing metadata inserted in the “Picture MetadataSet” area of the system item shown in FIG.
A “p” area, a “Picture Coding” area, and an “MPEG User Bitmap” area are shown. Further, the MPEG-2 picture editing metadata includes a “Profile / Level” area indicating the profile and level of MPEG-2, and an SMPTE1
It is also conceivable to provide video index information defined in 86-1995.

The one-word "Picture Edit"
ng Bitmap ”bits b7 and b6 are“ Ed
It flag ", which is a flag indicating editing point information. The following four types of states are indicated by the 2-bit flag.

00: No editing 01: The editing point is before the picture unit with this flag (Pre-picture edit) 10: The editing point is after the picture unit with this flag (Post-picture edit) -picture edit) 11: Only one picture unit is inserted, and the edit point is before and after the picture unit with this flag (single
le frame picture)

That is, a flag indicating whether the video data (in picture units) inserted into the picture item is before the edit point, after the edit point, or sandwiched between two edit points is set to “Picture”. Metadata
Set ”(see FIG. 4),“ Picture Edit ”
ing Bitmap "region.

Bits b5 and b4 are set to "Error"
flag ". This "Error flag"
Whether the picture contains an uncorrectable error, the picture contains a concealment error, the picture contains no errors,
Furthermore, it indicates whether it is in an unknown state. Bit b3 is "P
Picture Coding ”is the“ Picture
This is a flag indicating whether or not it is in the “MetadataSet” area. Here, when “1” is set, “P”
icture coding ”is included.

Bit b2 is “Profile / Lev”
"el". here,
When it is set to “1”, the “Metadata
The “Block” includes “Profile / Level”. This “Profile / Level” is M
MP @ ML or HP indicating PEG profile or level
＠ Indicates HL or the like.

The bit b1 is a flag indicating whether or not "HV Size" exists. Here, when “1” is set, “HV Size” is included in the “Metadata Block”. Bit b0 is
This flag indicates whether or not “MPEG User Bitmap” is present. Here, when “1” is set, “MPE” is added to the “Metadata Block”.
G User Bitmap "is included.

One word "Picture Coden"
“Closed GOP” is provided in bit b7 of “g”. This “Closed GOP” indicates that the GOP (Group Of Picture) when MPEG-compressed is Closed.
d Indicates whether this is a GOP.

Bit b6 contains “Broken Lin”.
k ”is provided. This "Broken Link"
Is a flag used for reproduction control on the decoder side.
That is, each picture of MPEG is arranged like a B picture, a B picture, an I picture, etc., but when there is an editing point and a completely different stream is connected, for example, the B picture of the stream after switching is There is a possibility that decoding is performed with reference to the P picture of the stream before switching. By setting this flag, it is possible to prevent the decoding as described above on the decoder side.

Bits b5 to b3 contain "Picture"
“Coding Type” is provided. This "Pi
"cure Coding Type" is a flag indicating whether the picture is an I picture, a B picture, or a P picture. Bits b2 to b0 are undefined areas (Reserved).

The one-word “MPEG User Bit”
"History dat" is set in bit b7 of "map".
a "is provided. This "History dat"
“a” indicates that the encoded data such as the quantization step, the macro type, and the motion vector, which are necessary for the encoding of the previous generation, are stored in the user data area existing in “Metadata” of “Metadata Block”, for example. , History data, or not. Bit b6 contains “Anc
data ”is provided. This "Anc dat
"a" is the data inserted in the ancillary area (for example,
This is a flag indicating whether or not data necessary for MPEG compression is inserted as Anc data in the user data area.

Bit b5 contains "Video inde
x "is provided. This "Video inde
“x” is a video index in the video index area.
This is a flag indicating whether index information is inserted. The video index information is inserted into a 15-byte video index area. In this case, the insertion position is determined for each of the five classes (1.1, 1.2, 1.3, 1.4, and 1.5). For example, the video index information of the 1.1 class is inserted into the first three bytes.

Bit b4 contains “Picture or
der "is provided. This "Picture o
“rder” is a flag indicating whether or not the order of each picture of the MPEG stream has been changed. In addition, MPE
Changing the order of the pictures in the G stream is required for multiplexing.

Bits b3 and b2 contain "Timecode
e2 "and" Timecode1 ". This "Timecode2", "Timecode1"
Is in the area of Timecode2, 1
ical Interval Time Code), LTC (Longitudinal Ti)
me Code) is inserted. Bits b1 and b0 include “H-Phase” and “V
-Phase "is provided. This "H-Phas
"e" and "V-Phase" are flags indicating which horizontal pixel and vertical line are being encoded at the time of encoding, that is, whether or not the information of the frame actually used is in the user data area.

Next, the audio item will be described. The audio item "Element Dat
“a” is “Element Dat” as shown in FIG.
"a" is "Element Header", "Audio"
“Sample Count”, “Stream Val”
id Flags "and" Data Area ".

The one-word “Element Head”
The bit b7 of “r” is “FVUCPVValid Fla.
g), and FVUCP defined in the AES-3 format standardized by the AES (Audio Engineering Society) is the AES of “Data Area”.
-3 format audio data (audio data)
Indicates whether or not it has been set. Bits b6 to b3
Is an undefined area (Reserved), and bits b2 to b0
Sequence number of 5-frame sequence (5-sequ
ence counter) is indicated.

Here, the five-frame sequence will be described. One frame is composed of 525 scanning lines (30/1.
001) synchronized with the video signal of frame / second,
When an audio signal having a sampling frequency of 48 kHz is divided into blocks of each frame of a video signal,
The number of samples per video frame is 1601.6 samples / frame, which is not an integer value. For this reason,
A sequence in which two frames of 1601 samples are provided and three frames of 1602 samples are provided so that five frames have 8008 samples is called a five-frame sequence.

The 5-frame sequence is synchronized with the reference frame signal shown in FIG. 11A. For example, as shown in FIG. 11B, the frames of sequence numbers 1, 3, and 5 have 1602 samples, and the frames of sequence numbers 2 and 4 have 1601 samples. The sequence number is represented by bit b2
Ｂb0.

The two-word "Audio Sample"
"Count" includes bits c15 to c15 as shown in FIG.
This is a 16-bit counter in the range of 0 to 65535 using 0, and indicates the number of samples of each channel. Note that all channels have the same value in the element.

One-word “Stream Valid”
"Flags" indicates whether or not each stream of eight channels is valid. Here, when significant audio data is included in the channel, the bit corresponding to this channel is set to “1”, and otherwise, the bit is set to “0” and the bit is set to “1”. Only the audio data of the channel set to "" is transmitted.

"S2-s0" of "Data Area"
Is a data area for identifying each stream of eight channels. “F” indicates the start of a subframe.
“A23 to a0” are audio data, and “P,
"C, U, V" are channel status, user bit, V
The parity bits include parity bits, parity bits, and the like.

Next, metadata for an audio item will be described. Audio editing metadata (Aud
io Editing Metadata) is a combination of editing metadata, error metadata, and source coding metadata. As shown in FIG. 12, the audio editing metadata includes one word “Field / Frame fl”.
ags ", 1-word" Audio Editing "
Bitmap ", 1-word" CS Valid Bi
tmap "and" Channel Status "
Data ”.

The number of valid audio channels is determined by the “Stream Val” shown in FIG.
id Flags ". If the flag of “Stream Valid Flags” is set to “1”, “Audio E
"Diting Bitmap" becomes effective.

[0091] "Audio Editing Bitm"
“first editing flag” of “ap” is the first field, “Second editing flag”.
“lag” indicates information on the editing status in the second field, and indicates whether the editing point is before or after the field with this flag. "Error
“lag” indicates whether or not an error that cannot be corrected has occurred.

“CS Valid Bitmap” is
"Ch" of n (n = 6, 14, 18, or 22) bytes
"annule Status Data" header, which indicates which of the 24 channel status words is present in the data block. Here, "CS
Valid1 ”is“ Channel Status ”
This is a flag indicating whether or not there is data in 0 to 5 bytes of “Data”. "CS Valid2"-"CS
Valid4 ”is“ Channel Status ”
This is a flag indicating whether or not there is data in 6 to 13 bytes, 14 to 17 bytes, and 18 to 21 bytes of “Data”. In addition, "ChannelStatus
Data ”is set to 24 bytes, and 2
Depending on the data of the 22nd byte, whether or not there is data from 0 to 21 bytes is indicated.
The 3-byte data is used as a CRC from 0 to 22 bytes. In addition, "Fielded / Frame flag
The flag "s" indicates whether the audio data of eight channels is packed in frame units or field units.

A general-purpose data format (General Data
Format) is used to carry all freeform data types. However, this free form data type does not include special auxiliary element types such as IT nature (word processing, hypertext, etc.).

Since the SDTI-CP format is composed of the data streams described above,
When transmitting and receiving a data stream in the I-CP format, the above-described data transmission device 10 can be configured as shown in FIG.

Here, the interfaces 22 and 2 shown in FIG.
13 correspond to 202, 203, and 204 in FIG. 13, the memory management unit 30 corresponds to 210 in FIG. 13, and the formatter 50 corresponds to the SDTI core unit 222 in FIG. Further, the memory 40 corresponds to the SDRAM section 250 of FIG.
0 corresponds to the signal conversion unit 260 in FIG.

The data transmission apparatus 10 is configured to share transmission and reception, and will be described from the data transmission side for convenience of explanation.

A CPU (Central Processing Unit) 24 is provided in the CPU interface 201 of the data transmission apparatus 10.
0 is connected. The CPU interface 201 has a plurality of registers. The plurality of registers are sequentially designated by the address signal SAD from the CPU 240, and the data signal SDT is supplied to the designated register to set “System Item Type” or “System
em Item Bitmap "FEC Activ
e Flag, SMPTE Label and RefDat
e / Time, Current Date / Time, and a flag indicating the presence or absence of Control;
nt Package Rate ”to“ Curren
The data up to “t Date / Time stamp” is stored. Further, these registers are provided in SDT described later.
Connected to I core section 222.

Further, the CPU interface 201 includes a memory control unit 211 of the memory management unit 210.
Is provided. This register is a storage location of data to be written to an SDRAM unit 250 described later, and is stored in “Package MetaDa”.
ta Set ”to“ Control Element ”
When the data up to "t" is stored in this predetermined register,
The stored data is sequentially written to the SDRAM unit 250 via the memory control unit 211 and the like.

The compressed video data DVC is supplied to the memory control unit 211 via the interface 202. The uncompressed audio data DAU is supplied to the memory control unit 211 via the interface 203. When the AUX data DSX, which is other data, is supplied, it is supplied to the memory control unit 211 via the interface 204.

Here, the interfaces 202, 203,
At 204, the number of words of the input data is counted in frame units, and not only the data but also the respective word count values VC, AC, and XC are supplied to the memory control unit 211. In the interfaces 202, 203, and 204, a clock change process is performed, and data input from the outside at different frequencies is stored in an SDRAM unit 2 described later.
Replace with data of 50 clock frequency.

For example, 27 MHz, 36 MHz or 5 MHz
8-bit video data DVC having a transmission rate of 4 MHz and 16-bit audio data DAU having a sampling frequency of 48 kHz are written in the memory 80, and the SDRAM
The clock is read out at 81 MHz or 108 MHz, which is the frequency of the clock signal of the unit 250, and the clock is switched.

Further, the interface 203 performs serial-parallel conversion to convert serial audio data into parallel data having a bit number corresponding to the bus width of the SDRAM section 250 and supplies the parallel data to the memory control section 211. The interface 202, 204 also converts the input data into a bit number corresponding to the bus width of the SDRAM unit 250 and supplies the converted data to the memory control unit 211. Similarly, in the CPU interface 201 described above, the SDRAM unit 250
In addition to counting the number of words for one frame of data to be written to the memory control unit 211 and supplying a word count value MC to the memory control unit 211, a clock switching process and a conversion process to a bit number according to a bus width are performed.

The memory control unit 211 includes the SDRAM unit 2
50 to store data to be written,
Is connected to a memory unit (first internal memory) 212 for temporarily storing data read out from the memory.

Here, the SDRAM section 250 is, for example, 32
When data can be written or read with the maximum transfer efficiency by performing burst transfer using 16 bits of bits, the memory unit 212 stores 32 bits for each of the system item, picture item, audio item, and AUX item. It is configured to have a memory capacity of 2 banks with 16 words of bits.

By configuring the memory unit 212 in this manner, by performing data transmission while switching banks, data of each item can be burst-transferred in 32 bits and 16 words.

In the memory control unit 211, the SDRAM unit 2
Memory unit 21 configured for data transfer with
2 is a single system, the arbitration of the writing of the data DSY, the video data DVC, the audio data DAU and the AUX data DSX of the system item supplied via the interfaces 201 to 204 is performed, and each data is sequentially stored in the memory. The information is stored in the unit 212. Thereafter, when data for the burst transfer is stored in the memory unit 212, a write request signal WQ for writing the data to the SDRAM unit 250 is generated and supplied to the arbiter unit 213. Further, the interface 20
The word count values MC, VC, AC, and XC supplied from 1, 202, 203, and 204 are stored in the word count table unit 214 for each frame.

In the arbiter 213, the memory controller 21
SDRAM unit 2 based on write request signal WQ from 1
SDRAM based on a write request to the memory 50 or a read request signal RQ from a memory control unit 215 described later.
The read request to the unit 250 is arbitrated. here,
By supplying signals AKA and AKB indicating that the request has been received to the memory control units 211 and 215, data transfer from the memory unit 212 to the SDRAM unit 250 and S
DRAM unit 250 to memory unit (second internal memory) 2
The transfer of data to 16 is performed at different timings.

Also, based on the word count value stored in word count table section 214, arbiter section 213 generates a control signal for writing and reading data to and from SDRAM section 250 and generates an SD signal.
By supplying the data to the RAM control unit 220, the data of each item can be correctly written to the SDRAM unit 250, and the written data can be correctly read.

The SDRAM control section 220 writes and reads data to and from the SDRAM section 250 by burst transfer based on a control signal from the arbiter section 213, and performs processing such as a refresh operation of the SDRAM section 250.

The SDRAM control section 220 further stores the end address of one frame of compressed video data. From this address, it can be determined that the data following thereafter is invalid data “0”.

The arbiter section 213 has an SDRA
A memory unit 216 for temporarily storing data read from the M unit 250 by burst transfer is connected. This memory unit 216 is, like the memory unit 212, an SDRAM
It is configured to be able to write and read data with the maximum transfer efficiency with the unit 250.

The writing and reading of data in the memory unit 216 is controlled by the memory control unit 215.
Data is written at the clock frequency of the SDRAM unit 250, the written data is converted into serial data, read out so as to be output at a transmission frequency (for example, 270 Mbps), and supplied to the SDTI core unit 222.

In the memory control unit 215, the memory unit 216
The data amount stored in the SDRAM unit 250 is determined, and a read request signal RQ for reading data from the SDRAM unit 250 is supplied to the arbiter unit 213 when the data amount decreases. If data exists in the memory unit 216, a request signal RQ for causing the data in the memory unit 216 to be read out by the SDTI core unit 222 is generated, and
Feed to 2.

In the SDTI core unit 222, the memory unit 21
The 8-bit item data read from 6 to 32 bits in 1-word units is converted into 10 bits. Also, CP
A system item is generated using information stored in a register of the U interface 201, or a picture item or an audio item is generated by adding item information or the like specific to a content package to compressed video data or audio data. . In the generation of each item, a CRC for error detection and an FEC for error detection and correction are also added.

Further, in the SDTI core section 222, “S
In addition to generating and adding fixed values such as "eparator" and "End Code", EAV, SAV and header data are added. The data generated in this manner is converted into serial data by the signal converter 230 and output, thereby transmitting serial digital data.

In the present invention, the following processing is performed to add the above-mentioned end code. First, FIG.
As shown in the figure, a video data storage area (picture item area) is provided in a part of the SDRAM section 250 which is an external memory.
Is provided. This video data storage area is an area to be written in frame units (or field units), and storage areas for a plurality of frames (4 to 8 frames) are prepared. In FIG. 14, video data storage areas (FL1 to FL4) for four frames are provided. The data length (the number of words) of each video data storage area is fixed and the same data length.

The video data storage areas FL1 to FL1
Invalid data (stuffing data) "0" is written in advance in all of FL4, and one frame of compressed video data including a header is written (overwritten) from the head area to corresponding storage areas FL1 to FL4. .

The compressed video data is variable-length data that differs depending on the video content of the frame. When the writing of the compressed video data is completed, the data “0” remains as it is because the remaining video data storage area has not been overwritten. The remaining data “0” is the stuffing data.

The amount of stuffing data is frame (FL)
1 to FL4), the stuffing data completely remains up to the last area of the video data storage area as shown in FIG. In FIG. 14, “x” indicates compressed video data that is effective data.

When reading compressed video data from the SDRAM section 250, all data in the storage area is read from the beginning of the video data storage area. In other words, only a predetermined number of words (including stuffing data) are read. By this reading, the variable-length compressed video data becomes fixed-length compressed video data. All data (including stuffing data) is read out, and when the number of words for one frame from the word count unit 214 is reached, the end code generated by the SDTI core unit 222 is added to this last data. Since the number of words to be read is constant in any frame, the management of the number of words to be read is easy, and the data management for adding an end code in the memory management unit 210 is reduced.

In the case where the data transmission apparatus 10 shown in FIG. 13 is used on the receiving side, a process reverse to that of the above-described transmitting side is performed. That is, as described above, the serial signal is converted into a 10-bit parallel signal by the signal conversion unit 260 and supplied to the SDTI core unit 222. SDTI core 2
In step 22, error detection is performed using a CRC or FE
C performs error detection and correction. Further, the data of each item of the content package is separated and stored in the memory unit 21.
Write to 6.

At this time, if "0" data is continuously detected for a predetermined number of words or more, for example, four words or more, the item data is determined to be stuffing data. Therefore, the stuffing data is stored in the memory unit 2
16 is not written. As a result of the fixed-length compressed video data transmitted by this process, the variable-length compressed video data itself, which is the effective data, is written into the memory unit 216. Also, the word count value included in the content package is extracted and stored in the word counter table unit 214.

The memory section 216 is provided with a memory control section 215
And the memory control unit 215 determines the amount of data written in the memory unit 216. Here, when the data for the burst transfer to the SDRAM unit 250 is stored in the memory unit 216, the memory unit 216
A write request signal WQ for writing the data stored in the SDRAM unit 250 is supplied from the memory control unit 215 to the arbiter unit 213.

The arbiter unit 213 includes the memory control unit 2
In addition to arbitrating the write request from the memory control unit 211 and the read request from the memory control unit 211, using the information on the number of words stored in the word counter table unit 214,
Video data and audio data for one frame of each item are transferred to the SDRAM unit 250 via the memory unit 216.
Write to. The data read from the SDRAM unit 250 is supplied to the memory unit 212.

The data written in the memory unit 212 is
The data is sequentially read by the memory control unit 211 and supplied to the interface units 201 to 204 corresponding to the data. In each of the interface units 201 to 204, the clock is switched by a process reverse to that at the time of transmission to perform SDRAM.
The data based on the clock frequency of the unit 250 is output as data of the clock frequency corresponding to the data recording / reproducing device 10.

[0126]

As described above, according to the present invention, variable-length compressed video data is treated as fixed-length compressed video data and an end code is added. Therefore, stuffing data is added to the compressed video data to obtain fixed-length compressed video data.

In this case, the data length of the compressed video data to which the stuffing data is added is always constant in any frame, so that the management of the compressed video data at the time of adding the end code is greatly simplified. Have.

Therefore, the present invention is suitable to be applied to a case where a transmission packet of the SDTI-CP format which handles variable-length data and a code indicating a delimiter thereof is transmitted, such as a data transmission system in a station. .

[Brief description of the drawings]

FIG. 1 is a configuration diagram showing a mapping example of an SDTI-CP format to which the present invention can be applied.

FIG. 2 is a configuration diagram showing a format of an EAV and header data.

FIG. 3 is a configuration diagram showing a format of a variable-length block.

FIG. 4 is a configuration diagram of a system item.

FIG. 5 is a configuration diagram of a time code.

FIG. 6 is a configuration diagram of a metadata set.

FIG. 7 is a configuration diagram of items excluding system items.

FIG. 8 is a configuration diagram of an SDTI-CP element frame.

FIG. 9 is a configuration diagram of MPEG-2 picture editing metadata.

FIG. 10 is a configuration diagram of an element data block of an audio item.

FIG. 11 is a timing chart showing a 5-frame sequence.

FIG. 12 is a configuration diagram of audio editing metadata.

FIG. 13 is a system diagram of a main part showing one embodiment of a data transmission device according to the present invention.

FIG. 14 is a diagram showing a data storage area of the SDRAM unit.

FIG. 15 is a system diagram showing an outline of a conventional data transmission device.

[Explanation of symbols]

10 data transmission device, 22, 24, 26, 20
2,203,204 interface, 30,210
... Memory management unit, 40, 250 ... S
DRAM, 50, 222 ... formatter (SDTI)
Core part), 60, 260 ... signal conversion part, 70 ...
Clock generator, 213 ... Arbiter part

 ──────────────────────────────────────────────────続 き Continued on the front page (72) Inventor Ichiro Tanji 6-7-35 Kita-Shinagawa, Shinagawa-ku, Tokyo Inside Sony Corporation (72) Inventor Hiroshi Higuchi 6-35, Kita-Shinagawa, Shinagawa-ku, Tokyo Sony Corporation F term (reference) 5C059 KK00 KK36 MC38 ME01 ME13 RB02 RB14 RC00 RC02 RC09 RC16 RC24 RF02 RF05 UA36 UA38

Claims (5)

[Claims] 1. A method according to claim 1, further comprising: adding invalid data to the variable-length compressed video data for one frame or one field to generate fixed-length compressed data. A data generation method for fixed-length compressed video data, wherein a code indicating a delimiter of compressed video data for a field is inserted. 2. A method according to claim 1, wherein a plurality of fixed-length data storage areas are provided in the memory, and invalid data is written in advance in each of the data storage areas, and compressed video data in units of one frame or one field is transferred from the head of the data storage area. 2. The fixed-length compressed video data according to claim 1, wherein fixed-length compressed video data is generated by overwriting and reading out all data in the data storage area including the invalid data. Data generation method. 3. The method according to claim 1, wherein the invalid data is data of logic “0”. 4. A section of one line of a video frame includes an end synchronization code area in which an end synchronization code is inserted, an auxiliary data area in which auxiliary data is inserted, a start synchronization code area in which a start synchronization code is inserted, and A serial area composed of a payload area into which data including compressed video data is inserted.
A data transmission device for transmitting a transmission packet of a digital transfer interface, wherein compressed video data inserted into the payload portion is 1
A data transmission device, which is fixed-length compressed video data obtained by adding invalid data to variable-length compressed video data for a frame or one field. 5. A memory for storing fixed-length compressed video data by adding invalid data to variable-length compressed video data for one frame or one field. Data transmission equipment.